Information-seeking behavior of women in their path to an innovative alternate treatment for symptomatic uterine fibroids.

OBJECTIVES
The purpose of the study was to discover how women found out about uterine fibroid embolization (UFE) for the treatment of symptomatic uterine fibroids.


METHODS
The study retrospectively tracked women from the beginning of their illnesses and analyzed their information-seeking behaviors. Twenty-eight women who had the procedure at the Detroit Medical Center were interviewed using a standard script. Median values and frequencies were calculated to represent information needs, information sources, and perceived helpfulness. Spearman correlations were calculated to find relationships between demographics and information needs.


RESULTS
Although the women expressed a great need for almost all types of information (median = 5), those with higher levels of education indicated a greater need to know the reasons that their doctors had for suggesting treatments (r = 0.55). The gathered frequencies indicated that friends, magazines, television, and the Internet were important information sources. A preliminary model of information seeking showed that while friends, magazines, and television made several of the women aware of the new procedure, the Internet was heavily utilized for learning about treatment options.


CONCLUSIONS
Health sciences librarians may inform women about their health and treatment options by guiding women to easily readable, authoritative, and reliable information sources, including Web information sites.


INTRODUCTION
Although women's health is a major discipline, women's health information needs and seeking is still a de-Supplemental tables and an appendix and figure are available with the online version of this journal.
veloping area of study. Relatively little has been researched and written about how women seek information pertaining to their health. In particular, research on women's information seeking in illness situations is scant, though it has been pursued in medicine, nursing, and allied disciplines, such as psychology [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17]. Most of the existing literature is focused on patients with breast or ovarian cancer [18,19]. Two themes, the types of information women need and the sources they use, have generally been approached separately for the most part, thereby leaving a gap in the literature for a probe on types of information that is received or sought from the various information sources.
The few studies in library and information science literature have dealt with the research problem differently and investigated the needs and behaviors of women not diagnosed with any particular illness [20,21]. Also, these studies tend to emphasize theoretical models [22,23]. Although this emphasis is expected in the library and information science literature, a gap remains in the literature regarding women's information behavior after diagnosis.
The present study, which is intended to be an important contribution to research on women's information needs in real health situations, probes how women who were diagnosed with symptomatic uterine fibroids found information about an innovative, minimally invasive procedure known as uterine fibroid embolization (UFE).
Uterine fibroids are nonmalignant tumors that occur in women of reproductive age, with the prevalence of these tumors increasing with age. Of women age 35 and older, 20% to 40% have uterine fibroids that are considered significant in size [24]. The traditional treatment for symptomatic uterine fibroids has been hysterectomy, an extensive surgical procedure performed by obstetrician-gynecologists [25]. In hysterectomy, the uterus is entirely removed surgically and the disease is completely cured. However, the removal of the uterus also means irreversible infertility for the women [25]. As a result, uterine-sparing treatments have become preferred alternatives for women with the disease. UFE is one such uterine-sparing, minimally invasive procedure and is performed by interventional radiologists.
The major benefits of UFE for women with fibroid disease, in general, are avoiding surgical risks and lengthy hospitalization. For women of childbearing age, there is also the potential for preserving fertility [25]. Thus far, the reports of clinical trials conducted to study UFE indicate that it is safe and effective in shrinking uterine fibroids and treating the symptoms of the disease in select patients [26][27][28][29][30][31][32][33]. Also, its efficacy in comparison to hysterectomy is promising [34].
Ravina and others in France first published results noting the success of UFE in 1995 [35]. While using embolization as an adjunct procedure to control bleeding before hysterectomy and myomectomy, another surgical but uterine-sparing procedure for treating fibroids, Ravina noticed that the embolization itself was having an effect of shrinking the fibroids. This discovery sparked a significant innovation in interventional radiology. Since 1995, interventional radiologists have integrated the practice of UFE. This author's study of interventional radiologists conducted in Michigan in 2002, however, indicated that radiologists still considered UFE a new procedure because its diffusion had been slower than expected [36]. While discussing issues involved in adopting the procedure, several radiologists mentioned that gynecologists, who are the primary or secondary contact for women and perform hysterectomy and other procedures for symptomatic fibroids, might be reticent to refer suitable patients to the radiologists for the new procedure. A later article in The Wall Street Journal reiterated the radiologists' concern [37].

OBJECTIVES
To further explore such issues, this study was designed to examine the information needs of women with symptomatic uterine fibroids who had been treated by UFE and the sources from which they received and/or sought information. This study explored the path through which women found out about the less invasive new procedure, retrospectively tracking women from diagnosis to procedure in order to understand the information sources that provided much-needed information in their path to UFE. It examines women's information needs, the ways women moved through the process of identifying a treatment, the kinds of information sources they used, and their evaluation of utility.

METHODOLOGY
Patients who underwent UFE between January of 2000 and March of 2005 at the Detroit Medical Center (DMC), a hospital affiliated with Wayne State University in Michigan, were sent letters about the study and asked to participate. In the letter, the patient was given an opportunity to refuse participation by leaving a message for the researcher that she did not wish to be contacted any further regarding the study. The protocol, which was developed in collaboration with the interventional radiology department at Wayne State University, was approved by the university's human subjects committee. As the study is retrospective and analyzes the behavior of women who had the procedure between 2000 and 2005 and as the information search behaviors of some of these women had commenced before 2000, the embolization procedure is referred to as innovative in this article.
Of the forty-nine women who had the procedure at DMC during the aforementioned period and were sent letters, one patient left the researcher a message expressing her wish to not be contacted any further. A week after sending the initial letters, the researcher called the participants to conduct telephone interviews. At this time, the researcher discovered that one patient was deceased, one did not speak English, three had uterine artery embolization for indications other than uterine fibroids, and nine had moved or could not be reached for various reasons over the telephone for an interview.
Overall, of the thirty-five women who had the procedure at DMC and for whom the researcher had accurate contact information, twenty-eight women participated in telephone interviews. Therefore, the response rate (80%) was high. Also, a majority of the women who participated expressed great interest in the study and were enthusiastic about their participation. Upon completion of the interviews, one participant's responses were construed as unlikely because all of her responses to questions with a five-point Likert scale were the highest and all of her responses to questions about information received or sought from various sources were nondiscriminatory; that is, all sources were noted as providing information. The participant was determined to be an outlier and was removed from the analysis [38].

Interview guide
The interview guide designed for the study (Appendix and supplemental material available online) was based on an instrument employed by Galloway et al. [10] in their study of the information needs of breast cancer patients, which also explored reliability and validity of the original instrument. Permission was obtained from the publisher of the instrument for the adaptation. All of the questions from Galloway et al.'s instrument that could be applied in this study were listed and measured on the same five-point scale (ranging from ''not important'' to ''extremely important'') in the first section of the interview guide. For example, the questions on incision and possibility of death from Galloway et al.'s instrument were excluded because these questions are only relevant in more severe illnesses. Also, the first section of this interview guide was the foundation Note: Numbers represent frequencies of use of an information source in meeting an information need. Participants indicated use of more than one information source for a type of information that she needed.
for the entire interview. In the present study, the adapted interview guide was tested in mock interviews with non-patients to establish clarity and comprehension.
After the mock interviews, the interview guide adapted for the present study of women diagnosed with symptomatic uterine fibroids consisted of four parts: questions about (1) the interviewees' information needs, (2) information sources employed, (3) helpfulness of the information sources as viewed by the women, and (4) demographics.
The information needs segment of the interview guide asked the women about the importance that they attached to various types of information concerning their illness, treatment, care, and the new procedure. These information needs were closely aligned with the needs identified by Galloway et al. [10].
The second part of the interview guide listed the possible sources of information for the women in receiving and/or seeking the various types of information they needed, including physicians, personal contacts, popular media, and the Internet. These information sources were based on a broad review of the female cancer patients' information needs literature.
It is important to mention that the information needs were listed in the order in which the women were expected to need information based on a preconceived sequence of stages during illness. For the purposes of this study, these preconceived stages of progression during illness from a patient's perspective were awareness of options, knowledge of options, decision, and the procedure. The items in the questionnaire that were intended to glean responses in each stage were: awareness of options: questions 17 to 20, knowledge of options: questions 21 to 25, decision: questions 26 to 31, and procedure: questions 32 to 33.
The patients responded to the queries about information sources that provided each type of information women needed during the progressive stages of ill-ness. In addition, the patients were forthcoming with several comments that complemented these responses. These comments were transcribed carefully with a reference number to the relevant query; for instance, comments related to question 17 were transcribed with a reference number Q17. When the responses about information sources were synthesized with these supplementary comments about receiving information, a preliminary model of information source use in learning about and undergoing embolization emerged.
In the process of building this preliminary model, the frequencies of information source use gathered in each stage were examined in light of the additional comments concerning information source, such as details of specific magazines and their use in meeting an information need at a stage of progression during illness, to determine the use of an information source exclusively at a certain temporal stage for embolization-related information. For example, information on benefits and risks of embolization were obtained by women throughout the illness. However, to determine the use of a source for this information at the knowledge stage, after awareness of the procedure and before a decision was reached, the raw frequencies of information source use were examined in the context of the women's additional comments. Also, when responding to the query on receiving information on available treatments, which might not include embolization, at the awareness stage of options, women did not select television; however, they mentioned in comments that they heard about embolization on a television show. In this case, although Table 1 does not show a frequency for receiving information on treatments available through television, the preliminary model notes television as a source for creating awareness of embolization specifically. The purpose of presenting frequencies in Table 1 was to provide data concerning information source for a type of information at anytime throughout the patient experience, from awareness of options to undergoing embolization, whereas the preliminary model only constructs prevalence of use of an information source for embolization-related information specifically at each stage, from awareness to undergoing the procedure in the patient's experience.
The women's comments that were transcribed validated the preconceived stages of progression during illness, with a modification required in only one stage: the knowledge stage. The knowledge stage emerged as 2 phases: conscious exploration and self-education. This modification in the stage emerged because the women's comments while answering questions 21 to 23 indicated that they were uncertain and thus still exploring their options through interpersonal means, whereas their comments related to questions 24 and 25 indicated that they were educating themselves through various means by seeking information on aspects that they were aware of and had explored to improve clarity. Also, several additional comments from the patients while responding to questions about information sources led to the data concerning family practitioners, internists, and gynecologists reported in the online only tables.
Respondents were also asked to rate the helpfulness of information sources using a five-point Likert scale of helpfulness (ranging from ''not helpful'' to ''extremely helpful'') modeled after the measurement of helpfulness employed by Schapira et al. [39]. The last part of the interview guide, designed by this researcher, included questions on age, education, and race and/or ethnicity.
In addition, the women were free to provide any comments about their experiences from the time they first felt ill to the stage of having the procedure and its aftermath, including any comments on whether the embolization was a success or failure and how they were faring at the time of the interview.
The interview was designed to be fifteen minutes long; however, most participants chose to talk to the researcher about their experiences beyond the allocated time. The researcher employed a standard script to conduct the interviews to ensure consistency between respondents. The interview questions were read to the participant after a brief introduction to the study and a description of how the interview would be conducted.

Statistical analysis
Median values were calculated to find the degree of importance women attached to various types of information. The median was chosen as the statistic to express the level of information need because the importance women attached to the different types of information was measured on an ordinal five-point Likert scale. Next, Spearman correlation coefficients (P Ͻ 0.05) were used to explore whether the importance that the women attached to various types of information was associated with their age and/or education (both variables measured as ordinal data in the interview). Lastly, women's use of information sources in receiving and/or seeking information was represented in frequencies, and median values were used to summarize the ordinal data representing the helpfulness of each information source to women. In addition, the Wilcoxon test was applied to find if gynecologists differed from the radiologists in the helpfulness of information they provided to their patients. The threshold for statistical significance was set at P Ͻ 0.05.

Demographics
The demographics of the women who participated in the study are shown in Table 2. The mean age of the women who participated in the study was 45.3 years (range 20-58 years). The youngest woman in the study was 20 years old, an uncommon age for this illness. Although 17 women had graduated from college, 10 women lacked college degrees, including 4 women who never attended college. Of the women, 21 were African American, 5 were white, and 1 was of mixed race/ethnicity that included Asian heritage.

Information needs of women diagnosed with symptomatic uterine fibroids
Among the twenty-eight women, almost all expressed a great need for information on diagnosis, treatment, and self-care. The median values of the importance that women attached to the various types of information are shown in Table 3. As can be seen, the majority of the information needs were rated as extremely high (median values ϭ 5).
Deviating from this pattern, the women gave less importance to only three types of information: doctor's reasons for suggesting a treatment (median ϭ 3), costs associated with treatments (median ϭ 1), and details of how the procedure worked in treating symptomatic fibroids (median ϭ 4). As can be seen, women attached relatively less importance to reasons their doctors, in this situation often gynecologists, gave for suggesting treatments ( Table 3). The technique and the mode of action for the embolization procedure-that is, how the procedure is performed by radiologists and how it works to treat the problem-however, was only slightly less important compared to need for other types of  information, which were rated extremely high by the women as represented by median values of 5.
Correlations were calculated for the women's need for each type of information, their ages, and their educational levels to find if these demographics affected the women's need for information. The Spearman correlation coefficients presented in Table 4 indicated that the women with higher levels of education did want more of certain types of information: different treatments available for symptomatic fibroids (r ϭ 0.55), benefits of various treatments (r ϭ 0.55), risks associated with various treatments (r ϭ 0.55), physician's reasons for suggesting a particular treatment (r ϭ 0.55), and costs associated with treatments (r ϭ 0.44). These were all treatment-related information; thus, the significant positive correlations indicated that the women with higher levels of education wanted more information on many treatment-related issues, including the physician's reasons for suggesting a particular treatment (P Ͻ 0.05).
The reasons physicians provide for suggesting a treatment, in particular, is critical patient information in the illness process [40]. However, as noted earlier, it was found to be less important (median ϭ 3) in this study. An examination of the percentages of women in Table 3 also shows that fewer women (n ϭ 12) said they found the reasons physicians provided for selecting a treatment for them to be very or extremely important. The correlation between women with higher levels of education and their need to know about the reasons for physician treatment recommendations suggested women with higher levels of education found this information more valuable.
Most participants were in their mid-forties with little variance in age in the sample. As such, no significant correlations were observed between age and need for any information, except for the information on length of time required before normal day-to-day activities could be resumed after the procedure (r ϭ Ϫ0.44). The significant negative correlation indicates that as the women's age increased, they attached less importance to knowing about the length of time required before resuming day-to-day activities (P Ͻ 0.05); that is, older women found this type of information less important (Table 4).

Information sources used by women diagnosed with symptomatic uterine fibroids
In participant responses identifying information sources, frequencies indicated that most of the information participants needed was received primarily from radiologists. The Internet was a major source for information in addition to the gynecologists (Table 1). Popular magazines also were identified as an important source of information. Among lesser or rarely used information sources were other primary care physicians, family members, newspapers, and the radio (Table 1). Women were also asked to rate the helpfulness of an information source in providing information that they needed. Table 5 shows that the radiologists were viewed as the most helpful in providing information (median ϭ 5). The Internet with a median value of 4 ranked next after the radiologists as a helpful information source. Following the Internet in helpfulness were the gynecologists and other primary care physicians with median values of 2 and 1, respectively.
In addition, during analysis, radiologists were compared to gynecologists regarding their helpfulness as an information source. The results of the Wilcoxon test indicated that radiologists were viewed as significantly more helpful than gynecologists (P Ͻ 0.05). The women did not agree about the helpfulness of the other information sources.

Toward a model of information seeking in women's path to an innovative procedure
In terms of existing work on developing a theoretical basis for women's health information behavior, Warner and Procaccino [22] have applied Kuhlthau's information search process (ISP) model to analyze women's health information needs, search strategies, and use of information in everyday situations. According to Warner and Procaccino, results appeared to address the uncertainty stage of the ISP model, as there were conflicting responses regarding the facility of locating information, the usefulness of the information found, and whether or not the subjects' health questions were answered. [22] The preliminary model in this study does not explore women's cognitive processes in information seeking; rather it focuses on women's information needs and their use of communication channels in information seeking in non-life-threatening health situations as they find their way to an innovative procedure. The frequencies gathered concerning types of information the women received from various information sources (Table 1), supplemented by women's comments during the interview, were the basis for the preliminary model. In tandem, the types of information in the second section of the interview guide follow a temporal sequence of need for information in the process of being diagnosed with symptomatic fibroids and having the embolization procedure.
The five stages through which the women moved as they became aware of the new procedure, sought information, and eventually had the procedure are: (1) awareness; (2) conscious exploration of options; (3) self-education; (4) probing, discussion, and decision; and (5) procedure (Table 6 and Figure 1). The stages are mutually exclusive only in a temporal sense; for instance, as self-education can occur at the procedure stage as well, the self education stage in the model refers specifically to self-education that occurs in a temporal sequence before a decision is reached concerning treatment in the process of a patient moving from diagnosis to procedure. The information sources used by women at each stage, either actively or passively sought, are discussed here.
At the awareness stage, women heard of the procedure through a mix of mass and interpersonal sources. Interpersonal sources predominated, including friends who were often also colleagues, rather than physicians. Few participants reported that they found out about embolization from their gynecologists and fewer from their family members. Mass media sources, such as television and magazines, also were instrumental at this stage for some women. Women described how Table 6 Use of mass media versus interpersonal sources from awareness of embolization to undergoing embolization (n ϭ 27)

Mass Interpersonal
Awareness Popular magazines (6) Television (5) Newspapers (2) Friend (9) (two of the friends were staff in primary care physicians' offices) Family member (2) Gynecologist (11) Conscious exploration of options Gynecologist (7) Self-education Internet (14) Popular magazines (3) Newspaper (1) Probe, discussion, and decision Gynecologist (3) Radiologist (4) Friend (1) Procedure Internet (2) Radiologist (7) Note: Numbers represent frequencies of use of an information source in meeting embolization-related information needs at a particular stage in the process of information seeking. A patient may indicate, and often indicated, use of more than one information source.
they had stumbled on the procedure through a television show or a magazine article, mentioning specific shows-such as 20/20, Primetime Live, 48 Hours, and Today-when they spoke about television. Among popular magazines, they often mentioned Essence, a magazine geared toward African American women. Other magazines in which the women had come across articles on the procedures for symptomatic uterine fibroids included Prevention, Women' s Day, and First. Noteworthy here is that the women who did not become aware of UFE through their gynecologists had already consulted with a gynecologist but became aware of the procedure through other channels.
Once they became aware of the new procedure through a source other than their gynecologists, they often went back to their gynecologists to ask about it ( Table 6 and Figure 1). No other source was mentioned at this stage of conscious exploration of options. At the next stage, self-education, the women relied heavily on the Internet for active information seeking (Table 6 and Figure 1). They spoke in a compelling tone about the possibilities for information and communication allowed on the Internet. Some women also noted that they read articles in popular magazines that they came across in their everyday interactions at this stage. This form of information seeking was often passive in that the women did not make an effort to get the information, but they came across it in their everyday interactions. These women were educating themselves by reading the magazine articles because they were already aware of the procedure, an awareness that came about through any of the various means, when they came across the articles that they spoke of reading at the self-education stage.
Upon exploration and self-education, participants often returned to physicians for further confirmation and discussion before they reached a decision (Table 6 and Figure 1). A mix of interpersonal sources was evident at this point: gynecologists, radiologists contacted directly by the women, and, to a minimal extent, friends (Table 6). At the procedure stage, most of the information was obtained from radiologists as can be expected (Table 6 and Figure 1). A few respondents reported that they were still searching on the Internet around the time of the procedure ( Table 6).

Obstacles faced by women in their path to the innovative procedure
Major obstacles faced by women-derived from open comments throughout the interview and, in particular, the second part of the interview guide-were that primary care physicians often only directed patients to gynecologists (Table 7), and gynecologists often suggested hysterectomy as the treatment to women ( Table  8). Also of concern was the indication that among the cases who were readily referred to the radiologists by the gynecologists, some were unsuitable candidates for the procedure. Of the seven women who had an unsuccessful embolization resulting in complications and/or further surgery (Table 9), five were patients to whom the gynecologists had suggested embolization. These five patients were part of the group of only nine patients in the study to whom the gynecologists had suggested embolization without being prompted ( Table 8), and, as can be seen in Table 8, the patients often received a referral to embolization after asking about alternatives. At times, however, the women who received a referral for embolization from gynecologists who were not prompted had contacted more than one gynecologist. Among the gynecologists of these nine women, four were the first gynecologists contacted, four were the second gynecologists approached, and one was the third gynecologist contacted by the patient.

DISCUSSION AND CONCLUSIONS
The percentage of African Americans in the sample was only slightly higher than expected, as the demographics of Detroit indicate that 81.6% are African American, 12.3% are white, and the remaining are of other race or ethnicity. Due to the demographics of Detroit as well the prevalence of the illness among African Americans [24], race must be given a consideration in any application of the findings of the present study.
Participants in this study-diagnosed with non-lifethreatening, symptomatic uterine fibroids-indicated a great need for almost all types of information. This was a striking difference from patients of cancer, a se-vere, chronic illness, who have been found to have a need for certain information but not a uniform need for all types of information related to their disease [10,11,[13][14][15][16][17]. Most remarkable in this study was participants' heavy use of the Internet and positive perception of the helpfulness of information obtained over the Internet. Some cancer patients, in contrast, have shown a relative reluctance to exploit the Internet for information [8]. These varying levels of use of the Internet might be attributed to the differences in the nature of the illnesses, non-life threatening versus life threatening, in which the illness itself affects information use.
The findings, including the unexpected findings on the channels that made women aware of the new procedure (friends, magazines, and television) and the importance of education level in active and/or assertive information seeking by women for treatment-related information, have implications for health sciences librarians in hospitals as well as public libraries. Given the great need for information that participants indicated, it is incumbent upon health sciences librarians to create and establish women's health information systems that can educate women on a number of lifethreatening as well as less severe health conditions. These systems must provide information on diseases, prevalence, demographics, and other concerns. In doing so, the importance of magazine articles, which often appear in a simple language, must not be ignored as a source for creating awareness. Most importantly, a concerted effort should be made to encourage women to ask their physicians about their health concerns. As the current study has revealed, some women, especially those with lower levels of education, do not probe their physicians about the treatment choices the physicians make for them.
It is also possible that some of these women would not approach an information professional or an organized information facility, such as a library. Therefore, proactive information dissemination efforts outside the library, such as introducing community-based information campaigns involving child-care facilities and salons, may be necessary. These initiatives are particularly critical in light of participants' perception of the lack of helpfulness of information provided by family physicians and internists.
Limitations of the present study include a small sample; hence, the study requires caution in formulating broad generalizations to populations beyond the hospital and treatment intervention chosen for analysis. The results of the study, however, are relevant to other women's health issues of comparable illness severity. Although individuals may recollect health-related experiences far more clearly than many others, the retrospective nature in querying women's experiences calls for caution in the interpretation of results. It may be worthwhile to note that, in this study, the women certainly did not hesitate and were clear in their descriptions as they spoke of their recollected experiences.
Additionally, as the majority of the patients at DMC who had the procedure by March 2005 were included in this study, a pilot study was impossible to conduct. This lack of a pilot study necessitates further testing of the instrument with similar populations. Also, the sample included only the women who had UFE. As such, the reader must be advised of limitations in validity because studies that include women who were informed about UFE but did not pursue it and women who were never informed about UFE and had hysterectomies might produce different responses to certain questions. In understanding the preliminary model of information seeking, the reader must be mindful of the small sample size and the temporal sequence of use based on a retrospective analysis of women's responses.
Future directions for study include an in-depth qualitative analysis of information seeking by the women. Also, a comparison of information-seeking behaviors of women in health situations of differing severity would be beneficial. The survey instrument can likely be applied in investigations of other women's health conditions.
In summary, potential implications for health sciences libraries and librarians include the following. First, the study indicates that many women in health situations of comparable severity may have a great need for information; therefore, health sciences librarians must actively engage in helping women find information. The study also suggests that some women may hesitate to probe certain questions concerning their treatment, which may require health sciences librarians to encourage all patients to be analytical in their interactions with mass media as well as interpersonal information sources. Most importantly, health sciences librarians are well situated to provide guidance in accessing easy-to-read and authoritative information sources, particularly reliable Web information sites, given the preference women have shown for this medium in seeking information. Table 7 Internist/family physicians' suggestions and/or treatments for women (n ϭ 27)

First internist or family physician
Second internist or family physician

Third internist or family physician
Not involved in discussion 4 No clear decision 1 1 Directed to gynecologist without discussing options 12 1 Discussed options that did not include embolization 2 Treated and then directed to gynecologist after some time 1 Table 8 Gynecologists' suggestions and/or treatment for women (n ϭ 27) Note: Numbers represent frequencies of women who received a particular suggestion and/or treatment from their first, second, or third gynecologist.

Table 9
Success of the new procedure as viewed by women (n ϭ 27)

Successful embolization Failed embolization Not sure
Gynecologist suggested embolization without patient asking for options 2 (7.4%) 6 (22.2%) 1 (3.7%) Gynecologist suggested embolization after patient asked for options 6 (22.2%) 1 (3.7%) Patient suggested embolization and gynecologist supported 1 (3.7%) Patient decided by herself to have embolization (some of these women consulted the radiologist directly) 7 (25.9%) 2 (7.4%) Note: Numbers represent the frequencies of women who had a particular decision-making pattern prior to the procedure and a successful versus unsuccessful embolization later.

Figure 1
The role of information sources in creating awareness of the procedure and facilitating decision making about the procedure among the women at different stages of information seeking.